Vol. 1, No. 1, January 1962
MECHANISM OF LIVER ALCOHOL DEHYDROGENASE REACTION.
111
41
Studies on the Mechanism of Enzyme-Catalyzed Oxidation Reduction Reactions. 111. A Characterization of the Mechanism of the Liver Alcohol Dehydrogenase Reaction* R. H. BAKER,J R . t From the Department of Chemistry, $ Indiana University, Bloomington, Indiana Received August 11, 1 9 6 ~
The mechanism of the reaction catalyzed by liver alcohol dehydrogenase (LADH) has been investigated by studying its kinetics: (a) by the method of initial steady state velocities; (b) by determining the kinetic isotope effect on these rates produced by the stereospecific introduction of deuterium into DPNH and ethanol; (c) by determining the effect of product inhibition on these rates; and (d) by the method of continuous steady state kinetics. The data so obtained were treated by means of the digital computer program described previously. The most likely mechanism, on the basis of the Haldane, Dalziel, and product inhibition criteria, is one requiring the compulsory binding of the coenzyme as the first step in the reaction. This conclusion is confirmed and extended by the isotope effect and continuous rate studies. These suggest in addition the occurrence of one (or more) ternary complexes as a required intermediate. In the case of those oxidative enzymes which catalyze the actual transfer of hydrogen between reductant and oxidant a useful and powerful tool of physical organic chemistry has become available to a study of mechanisms, viz., the application of kinetic deuterium isotope effects. With deuterium substitution in appropriate positions of reduced substrate or coenzyme a kinetic isotope effect can be expected in all those steps of a particular mechanism where a carbon-hydrogen bond is loosened or broken (for reviews see Wiberg, 1955; Streitwieser, 1960; Melander, 1960). This technique has been applied to the elucidation of the mechanism of many organic and some enzymatic reactions’ (Rieder and Rose, 1959; Abeles et al., 1960; Rose, 1961) and has been used in our laboratory for a preliminary survey of the yeast alcohol dehydrogenase reaction (Mahler and Douglas, 1957; Shiner et al., 1960). In this paper we shall describe the application of this technique to a different enzymatic reaction, i.e., the one catalyzed by liver alcohol dehydrogenase. Two stereochemically distinct forms of the deuterium-containing reduced coenzymej2both prepared enzymatically, have been employed. From the first one, the “a-” or A-form, deuterium * The investigations reported here were supported, in part, by a grant [ONR 908( 12)] from the Office of Naval Research to Indiana University. t National Science Foundation Co-operative Predoctoral Fellow, 1959-60. Present addresa, Department of Chemistry, University of Wisconsin, Madison. The experimental work reported here is taken in large part from the disaerts tion of the author, submitted to the Graduate School of Indiana University in partial fulfillment of the requirementa for the Ph.D. Degree. $ Contribution No. 1029.
1 Additional but by no means complete citations may he found in the contributions by Shiner et al. (1960), Collins (1960), and Englard (1960) to a symposium on this topic (Ann. N . Y . Acad. Sci. 84,573-881, 1960). 2 The stereospecificity of the hydrogen transfer in pyridine-nucleotiderequiring dehydrogenase reactions is discussed by Vennesland and Westheimer (1954) and Vennesland (1958).
is transferred in the reaction catalyzed by liver alcohol dehydrogenase and hence a primary isotope effect may be expected in the transfer step proper. From the second, the isomeric “p-” or B-form, hydrogen is transferred while deuterium is retained in the oxidized coenzyme; here only a secondary isotope effect can be expected. In addition, other criteria for the selection of an appropriate mechanism, discussed in the previous paper (Baker and Mahler, 1962),have been employed. Horse liver alcohol dehydrogenase was chosen for the following reasons: (1) The enzyme is readily obtainable in highly purified and active form (Bonnichsen and Brink, 1958; Dalziel, 1958). (2) The reaction involves the reversible oxidation and reduction of pyridine nucleotides and can be followed conveniently and accurately by a variety of optical methods (Theorell et al., 1955; Boyer and Theorell, 1956; Theorell and Winer, 1959). (3) The reaction is stereospecific with respect to both the substrate and the coenzyme (Vennesland and Westheimer, 1954; Levy and Vennesland, 1957; Vennesland, 1958). (4) Largely as a result of comprehensive investigations in Theorell’s laboratory (Theorell, 1956, 1958), a great deal of information concerning the enzyme itself is available: (a) It is exceedingly stable on storage. (b) Its electrophoretic mobility, molecular weight, and UV extinction coefficienthave been determined with great accuracy (Ehrenberg and Dalziel, 1958). (c) The number of equivalent, active sites has been determined to be two (Theorell, 1956, 1958; Vallee and Coombs, 1959). (d) Extensive information has been gathered on the binding of the coenzymes (Theorell and Winer, 1959; Van Eys et al., 1958; Vallee et al., 1959), and some data have been obtained on the binding of the substrates (Vallee et al., 1959; Theorell and Winer, 1959). (e) The reaction catalyzed by the enzyme was postulated to obey the classical, simple Theorell-Chance (1951) mechanism, although in recent years there have been reports which cast some doubt on the
42
Biochemistry
R. H. BhKER, JR.
validity of this conclusion (Theorell, 1958; Theorell and Winer, 1959; Winer, 1958; Theorell, 1961). (f) The substrate specificity has been investigated (Winer, 1958; Merritt and Tompkins, 1959), but only by measurement of pseudomaximal velocities and not by detailed study of the kinetics of the over-all reaction.
EXPERIMENTAL
of buffer to make our standard stock s o l ~ t i o n . ~ This solution is stable and does not change in concentration over periods up to 5 hours when ' 4 stored at . D. ETmNoL.--Ethanol, 100% pure, was obtained from B. Murr of this department, and stock solutions were prepared by delivering 10 pl portions from a micropipet a t room temperature into 50 ml of buffer. E. DIDEUTEROETHANOL (CH&D20H).-The dideuteroethanol was prepared by the method of Shiner (1952); except that Diethyl Carbitol was used as solvent. The LiAlD4 (98.5 atom % D) used as the reducing agent for the conversion of acetic anhydride to the alcohol was obtained from Metal Hydrides, Inc., Beverley, Mass. The azmtropic mixture (95.5% ethanol 4.5% water by weight) was collected at 78.4' with a yield of 77.8%. The deuterium content was 1.969 0.013 atoms of deuterium per molecule, as determined by Shiner's density column method. F. ~ - ~ - D E u T E R o - D P N H . -procedure T~~ of Rafter and Colowick (1958) was modified for the use of 2 g of DPN+ and 2.3 g of dideuteroethanol. A 60% yield of the yellow solid a-D-DPNH was obtained. Using the molecular extinction coefficient for DPNH given above, it was found to be 82.7% pure by weight; 100% of this material was enzymatically active. For kinetic runs, the barium salt of a-D-DPNH was converted to the sodium salt by addition of Na2S04followed by centrifugation and decantation. Deuterium determination (density column) showed0.952 i 0.018 atoms of deuterium per molecule of product disodium salt. 0.
[email protected] @deuteratedcoenzyme was prepared by Dr. R. Suzue in our laboratory by reaction (1) catalyzed by the enzyme, liver glucose dehydrogenase. The deuterated co-
Enzyme.-The twice-recrystallized enzyme was shipped to us at dry ice temperatures by Worthington Biochemical Corporation, Freehold, N. J. This preparation when stored at -20' retained all its activity over a &month period. For storage of stock solutions, we recrystallized the enzyme two additional times, recovering 97% of the units present originally, with specific activities of 2 100% based on the assay of Bonnichsen and Brink (1958). This solution was approximately 5 X M with regard to enzyme. The concentration of ammonium sulfate was brought to 38% (38 g/100 ml buffer). The solution was readjusted to pH 7.5 by the addition of ammonium hydroxide. When stored at -1l0, it retained its full activity for more than 45 days. For kinetic runs, up to 5 mg (approximately 0.2 ml of the storage stock solution) of protein was placed on a 10-cc Sephadex G-50 columm (Pharmacia, Uppsala, Sweden) and eluted with 2.7 ml of 0.10 M, pH 7.0 phosphate buffer with no loss of activity. This secondary stock solution was approximately 1 X lov6 M with regard to protein and was stored at 0' with no loss of activity over the period of a kinetic run. Just prior to any such run the solution was assayed according to the method of Bonnichsen and Brink (1958); for our calculation we employed a value for the turnover number of 1.33 X 106A ODw/ min./mole of enzyme, the last value based on the ldeutero-j3-wglucose + DPH+ most recent values for the molecular weight and D-gluconolactone j3-D-DPNH ( 1 ) ODm of the enzyme (Ehrenberg and Dalziel, 1958). enzyme was found to be 8l.Oy0 pure by weight Reactants and 100% enzymatically active. For kinetic runs, A. DIPHOSPHOPYRIDINE NUCLEOTIDE, OXIDIZED the @-D-DPNHwas treated in a manner analogous FORM (DPN +) .-DPN + was obtained commercially to that described for the a-D-DPNH. The 8-D(Sigma Chemical Company, St. Louis, Mo.). DPNH was found to contain 0.946 0.021 atoms It was found to be 95.0% pure, by weight, on the of deuterium per molecule (density column). basis of its ODrn, using the molar extinction H. BuFFER.-The buffer used in all of the coefficient 1.80 X l o 4 cmZ/mole (Kornberg and initial rate runs was 0.10 M glycylglycine-sodium Pricer, 1953). Enzymatic assay in the presence of hydroxide, pH 8.58 f 0.02. It was prepared from excess ethanol at pH 9.6 showed it to be 92.5% C.P. glycylglycine obtained from the Aldrich Chemactive. ical Co., Milwaukee, Wis. All of the solutions B. DIPHOSPHOPYRIDINE NUCLEOTIDE, REDUCEDwere made up in doubly distilled water, the second FORM(DPNH).-DPNH was obtained commer- distillation using an all-glass apparatus. For the cially (Sigma Chemical Company). It was found to continuous rate runs the buffer was 0.10 M imidabe 89.7% pure by weight from its ODw, using the zole-HC1, pH 7.58, prepared in an analogous manner the molar extinction coeflicient 6.22 X loa cmz/ from C.P. imidazole obtained from the same mole (Horecker and Kornberg, 1948). It was source. All pH determinations were performed determined to be 100% enzymatically active on with a Beckman pH meter, Model GS. the basis of its oxidation in the presence of excess Kinetic Runs.-All of the kinetic runs were acetaldehyde at pH 7.5. followed spectrophotometrically in a Cary Model c. ACETALDEHYDE.-Acetaldehyde was ob- 1 1 spectrophotometer (Applied Physics Corporatained commercially and redistilled prior to each tion, Pasadena, Calif.) fitted with a specially use in a 4 ' cold room. The fraction boiling at 20.0 0.2O was collected, the first few ml of dis'The 5+1pipet was calibrated by weight of water tillate being discarded. With a micropipet, 0.005 delivered t o sealed vials at 4' to be 0.00500 d= 0.00034 g. ml of the acetaldehyde was delivered into 10 ml The density of metaid6hyde at 4' is 0.7996 g/cc.
+
*
+
*
*
VoZ. 1, No. 1, January 1962
MECHANISM OF LIVER ALCOHOL DEHYDROGENASE REACTION.
constructed thermostated cell carriage for the sample cuvet. The cell carriage, substrate solutions, and washing solutions were held at the same constant temDerature with a Wilkens-Anderson constant temperature bath set a t 27 f 0.02'. A tungsten lamp was employed, and measurements were made a t 340 mp. With the 0.1 absorbancy slide wire, the slit control was set a t 0.7-mm slit width to maintain a noise level less than f 0.0003 OD units. Under these conditions the resolution of the spectrophotometer is f 2 mp. TABLE I CONDITIONS FOR INITIAL RATERUNS Temperature 27' f 0.02, pH 8.58, ionic strength 0.10 (glycylglycine NaOH) __ - __ Acetaldehyde DPNH enzyme 8.46 X lo-* site equivalents/liter M acetaldehyde 1.17 X lo-&to 9.01 X DPNH. 8.95 X 10-7 to 6.93 X I n 4 M Ethanol DPN+ enzyme 2.82 X lo-' site equivalents/liter ethanol" 2.20 x 10-4 to 1.70 x 1 0 - a ~ DPN 1.52 X lO-'to 1.18 X lo-' M 4 Slightly higher concentrations, leading to initial rates of the same magnitude, were used when deukrium-containing reactants were employed.
+
2.18
I,
I100 IOOO[
43
111
x
10-4 M
n
0
? { >-
+
+
1
IO
12 x 104
2 4 6 8 The reaction vessels were 3-ml quartz cuvets with a light path of 1.00 cm. The cuvets were IADPN+), M-* filled with 1.5-ml Dortions of the reactants bv FIG.1.-A typical initial steady state rate run. means of 2-cc Luer-lok syringes (Becton, Dick&son and Company, Rutherford, N. J.) calibrated that some reiterative method would perhaps lend with a buffer solution to deliver 1.50 0.01 ml. itself better to a statistical analysis of the problem. After the cuvets were placed in their compart- A matrix of four concentrations of A and B was ment, the recording system was turned on, the chosen to be run in triplicate. I n this way, we enzyme solution was pipetted by means of a could not only observe the experimental variation 50-pl pipet into a specially designed lucite mixing for any particular value but also have the choice
*
TABLE I1 STATISTICS OF SUBSTRATE CONCENTRATION-VELOCITY MATRIX (per cent of triplicate sets used-per cent range between points used) Coenzyme (decreasing concentrations) + substrate (decreasing concentrations)
+
i
(90.2-4.2) (85.6-5.8) (86.3-5.4) (80.8-5.1)
(78.7-4.9) (81.1-5.9) (80.0-6.3) (80.0-5.8) (80.8-5.9) (74.7-5.3) (79.4-5.7) (74.2-5.8) Over-all average for the sets (80.2 3.4-5.53 f 0.44)
spoon, and the reaction was started by a few rapid up and down motions of the spoon inside the cuvet. Complete mixing was achieved and the spectrophotometer responded to the reaction within 3 seconds.' To obtain initial rates, the observed rates on the Cary graph were extrapolated to time zero by means of a Keuffel and Esser french curve, model 1864-60. The slopes of the resulting extrapolated lines were determined and employed as initial rates. The substrate concentrations used in the initial rate runs (Table I) varied over a range of 7.75 fold. As mentioned previously (Baker and Mahler, 1962), up to ten different concentrations of A and B can be treated by the IBM program. Although such a complete matrix containing 100 points would undoubtedly be exceedingly useful, we decided 4 The completeness of the mixing waa checked a t 280 mp by the introduction of concentrated enzyme solutions into the cuvet.
(83.3-5.3) (79.4-6.0) (76.0-6.3) (72.0-5.0)
*
of discarding any point in a set of three if it seemed to fall far outside the mean determined by the other two. Table I1 summarizes the statistics of the matrices used. The rates gathered at the lower concentrations of reactants appear to be slightly less reliable. However, this fact, if true, became apparent only after all the points had been collected. Our understanding of the errors involved did not allow us to make an a priori prediction, a prerequisite for most computations of weighting factors. This then suggests the use of the computed deviations of one particular kinetic run as weighting factors for this particular run. I n essence we have done this in the following way: After the data for a run were collected and the errors involved computed, the best values falling within the standard deviations were selected so as to give a linear response in our IBM program. This is represented in Figure 1. The open circles
44
Biochemistry
R. H. BAKER, JR.
Buffer, 0.1 M
TABLE I11 CONDITIONS FOR CONTINUOUS STEADYSTATE RATERUNS imidazole, pH 7.58; enzyme, 6.98 X lO-'site equivalents/liter: temperature, 27"
Acetaldehyde
+ DPNH e
Ethanol
Acetaldehyde DPNH a-D-DPNH
+ DPN+ e
Ethanol Dideuteroethanol DPN+
represent the points actually used on the IBM-650, while the bars represent the observed deviations. The conditions used for the continuous steady state runs are summarized in Table 111.
8.56 8.30 8.17
x x x
10-4~ 10-4 M 10-4~
2. Haldane Relations (Relations Between cplZf/cplz and CPOCPI'PZ'/~~O'PIPOZ or (PO'~~'PZ'/(POPIY~). Keq
X lo-?
( ~ 1 2 ' / ~ 2
'poopi'w'/'po'(~iw
Hydrogen 3.56 f 0.24 3 . 7 1 f 0.82
Deuterium 3.41 f 0.28 2.05 f 0.70
x 10-1 1.08 f 0.24 0.42 f 0.14 x 10-3 A comparison of the ratio (alz'/cplz with the other two ratios shown indicates that the first is not equal to either of the other two except possibly for C P O ' P ~ ' P Z ' / Pin O ~the ~ ~ (deuterium ~~ example. This criterion is sufficient to eliminate mechanisms Ib, IIc, and IId; it permits no choice between the allowed mechanisms Ia, IIa, and IIb. 3. Dalziel (1958) Relations (Relations Between P I P ~ / P and ~ Z POOT PO').
RESULTS
'po'pi'm'/quopi~
1. Initial Rate Kinetics.-The
experimental conditions used have been summarized in Table I , while the average cpp values obtained are tabulated in Table IV. Also recorded are the standard deviations of the actual values for sets of separate kinetic runs, and the mean of the computed standard errors for the sets. Since the theory of errors in enzyme kinetics has not yet been developed to the point of permitting a critical appraisal of values such as those of Table IV, we make no attempt to explain their magnitude. The relative magnitude of the two types of error estimates shown is, however, of some interest. In nearly all cases the mean of the standard errors, a measure of the reliability of the individual 'pp value, was equal to or greater than the mean of the deviations between runs, an approximate measure of their reproducibility. Therefore - especially since the deviations were calculated on the basis of only two or three runs, while the standard errors were computed from the large number of data involved in a single runwe have used only the latter statistic for our calculations. The results shown in Table I V were then subjected to the methods for mechanism characterization described previously (Baker and Mahler, 1962; see Table I V in that paper for numbered list of mechanisms).
( ~ i d o p t zX
10' ( ~ i ' ~ ' / ( p i i ' X 10' 'Po x 102 'Po' x 10'
Hydrogen 3.35 f 0 . 3 8 1.88 f 0.15 6.79f0.40 3.67 f 0.13
Deuterium 6.47 f 0.92 1.76 f 0.22 10.8 f 1 . 7 4.86 f 0.04
PIP ZIP^^ z PO and cpl'cp2'/cpl2' z PO' in all cases. Therefore, mechanism I b is excluded. (p1(p2/(p1? 6 PO' in the hydrogen case, but 3 PO' in the deuterium example; similarly cpl'(a2'/(a12 6 cpo in both the hydrogen and deuterium examples. This is the reverse of the Dalziel relation, and is not exhibited by any of the mechanisms proposed by him; the magnitude of this effect, particularly in the case of the data involving ethanol plus DPN, is significant, and suggests that the actual mechanism must be a modifid Type I1 mechanism. 4. Product Inhibition.-Our studies of product inhibition were not too fruitful' owing to the dif-
TABLE IV SUMMARY OF RESULTSFOR INITIAL RATERUNS Reactants
Acetaldehyde DPNH Ethanol DPN
Number of Experi-
+
+
cp,
3
6.79 X -. 2.85 x e 1.41 x $012 1.20 x 'pol 3.67 X 01' 4.21 X (DDll 1.91 x ,plzt 4.27 x
2
+
Acetaldehyde a-I)-DPNH
+
2
'po
cpo (02
opir
Dideuteroethanol DPX+
+
2
cpo' 91'
op2r
+
Acetaldehyde @-D-DPNH
Values
mentsa
opf2'
2
@,
1.08 X
4.63 x 4.02 x 2.87 X 4.86 X 4.36 X 3.94 x 9.79 x 1.00
x
2.94 x 9 1.41 X VIP 2.77 x a Four concentrations each of A and B, each set run in4xiplicate. sec. M/site; and cp12, sec. M*/site. cpl
Standard Deviation (Experimental)
(%I 3.1 12. 10-6 6.8 10-11 2.8 lo-' 11. 1.7 10-4 1.6 10-9 0.12 10-1 20. 10-7 6.5 lo-' 0.12 10-1' 1.7 IO-' 1.8 lo-* 6.5 10-4 1.3 10-9 4.3 10-1 18. 10-7 5.5 lo-' 3.6 10-11 6.9 The units of 9 are: e, sec./site;
Computed Mean of Standard Errors
1%) 6.0
lo-* 10-7
cpl,
3.5 2.3 5.6 4.3 4.3 2.6 1.3 16. 8.5 0.97 3.0 0.84 5.0 2.4 5.2 13. 5.5 2.3 2.3 sec.M/site;
(cz,
Vol. 1! No. 1, January 1962
MECHANISM OF LIVER ALCOHOL DEHYDROGENASE REACTION.
111
ficulties inherent in the required alteration of substrate concentrations to obtain suitable values for velocity measurements. Some of the results and our conclusions are tabulated in Table V. Although the data are not completely unambiguous, largely because of the rather large errors involved in the determination of the A values, it is clear that the bulk of the evidence suggests one of the ordered sequence (compulsory binding) Type I1 mechanisms, with the coenzyme as the leading substrate. The ambiguities shown might be overcome, perhaps, by some modification of the Type I1 mechanism not previously considered in the derivation (Alberty, 1958; Baker and Mahler, 1962). 5. Isotope Eflects (Baker and Mahler, 196.2).The following relationships may be computed from the data of Table IV, excluding for the moment the data for fi-D-DPNH (i.e., the isomer from which deuterium cannot be removed in this reaction).
45
is satisfied, but $0 # +2 and +1 # q12, and mechanisms Ia and I b are therefore excluded. Since pl' equals unity, I a or I b are again impossible, but all other mechanisms are possible. Since go # 1 and probably &' # +1+2/+12, mechanisms IIc and d are held to be less likely than mechanisms IIa or IIb. 6. ,ContinuousSteady State Kinetics.-Investigations of continuous steady state kinetics were run under the conditions summarized in Table 111. These four continuous rate experiments were used to provide over 100 separate observations on concentrations and their rate of change, each determined in triplicate. These were placed on the computer, which then calculated some 300 rates and their respective errors, providing the data for a test of equation (23) of Baker and Mahler (1962) and a computation of its coefficients. For brevity, only one of the actual experiments will be shown (Table VI). All four are summarized in Table VII. The substrates employed for the $0 = 1.59f 0.36 $0' = 1.32f 0.06 data of Table VI1 were (1) dideuteroethanol, $1' = 1.04f 0.10 $1 = 1.62f 0.19 $2' = 2.06f 0.10 $2 = 2.84f 0.09 (2) ethanol, (3) a-D-DPNH, and (4) DPNH. $12' 2.29 f 0.15 $12 = 2.39 f 0.20 The errors given are the observed standard errors $1$2/*2 1.93i 0.46 calculated for each point taken in triplicate, the The identity +2' =gI2'is common to all mechanisms standard errors computed from the closeness of fit, and is satisfied by the data. The identity 30' =$1' and the average absolute fitting errors (positive TABLEV SUMMARY OF RESULTS FOR SUBSTRATE INHIBITION RUNS (Dp
Acetaldehyde
+ DPNH +ethanol (5.65X lo-' M )
'Po
VI M
VU
Acetaldehyde
+
,.
@I,
91
i.56 X
M VI2
Ethanol
8.24x lo-* f 0.79 2.40X lo-' f 0.04 2.39 X lo-' f 0.06 2.26X lo-" f 0.05 DPNH +DPN+ (2.51X lou5M ) 1.15X 10-l f 0.16 3.31 X 10-7 f 0.22
+
I
io-6
f 0.02
5.37x 10-11 f 0.04 DPN+ +DPNH (9.82X lo-' M) 3.98X lo-' f 0.12 8.07X 10-5 f 0.21 1.29 X lo-' f 0.05 3.20X 1O-O f 0.13
Acetaldehyde
+ a-D-DPNH
----f
+
+
+
VI,
Indeterminate; some support for I
1.08f 0.07 1.92 f 0.13 .~ . ~0 . 6 7 i 0.04 0.75i 0.04
Indeterminate
sec. M / S i k ;
1.54 f 0.28 0.45f 0.09 1.08f 0.02 1.87f 0.09
Indeterminate
0.82f 0.19 4.60f 0.38 1.30f 0.02 6.40f 0.23
Probably I1
1.06f 0.02 3.20f 0.21 0.38f 0.03 2.34f 0.15
Probably I1
1.11 i 0.12
Probably I1
M)
7 "
are: a,sec./site;
1.68f 0.33 1.16i 0.12 0.90f 0.04 4.47f 0.28
dideuteroethanol(5.45 X lo-' M)
1.67x 10-1 f 0.03 2.09 X 10-7 f 0.23 VI 4.38X 10-6 f 0.03 'h 5.39x 10-11 f 0.10 VlZ Acetaldehyde a-D-DPNH +DPN + (2.42X m 8.81X lo-% f 0.63 i.85X 10-6 f 0.05 Vl 5.25x 10-6 f 0.02 M 1.84X 10-10 f 0.01 VlZ Dideuteroethanol DPN+ +a-D-DPNH (1.63X 5.18x 10-1 f 0.05 'Po' M' 1.40 x 10-5 f 0.02 M' 1.51x 10-4 f 0.09 (012' 2.29x 10-6 f 0.03 Ethanol DPN+ +b-D-DPNH (3.06x M) 'Po' 4.09X 10-1 f 0.29 M' 1.20 x 10-6 f 0.04 (4' 1.79x 10-4 f 0.10 (42' 1.08x 10-8 f 0.01 'p
Probably I1 (d) ?
~
m ,-
The units of
1.21f 0.19 0.84f 0.05 1.69f 0.08 1.88 f 0.15
M)
2.84 f 0.21
0.94f 0.08 2.53 f 0.06 e, sec. M/site; and .m2,sec. Mz/site.
46
Biochemistry
R. H. BAKER, JR.
TABLEVI most likely mechanism for the reaction catalyzed DATA FROM EXPERIMENT O N CONTINUOUS STEADY-STATEby liver alcohol dehydrogenase is a steady-state KINETICS one involving a compulsory order of binding of Reaction: DPN dideuteroethanol (continuous rate reactants, with the coenzyme bound first in an runs) +
Observed Rates, Site/Seo.
0.1205 0.1033 0.0910 0.0825 0.0756 0.0710 0.0637 0.0580 0.0495
0.0430 0.0369 0.0311 0.0253 0.0207 0.0169 0.0146 0.0123 0.0100 0.0096 0.0088 0.0069 0.0054 0.0042 0.0031 0.0023 0.0019
+
All Constants
Computed Rates
bi
bl and br
=o
0
0.1190 0.1163 0.1035 0.1049 0.0925 0.0944 0.0836 0.0849 0.0760 0.0764 0.0686 0.0863 0.0622 0.0615 0.0562 0.0553 0.0504 .~~~ 0.0495 0.0457 0.0449 0.0376 0.0372 0.0308 0.0308 0.0252 0.0255 0.0210 0.0214 0.0174 0.0180 0.0148 0.0154 0.0124 0.013C 0.0110 0.0105 0.0093 0.0089 0.0075 0.0079 0.0063 0.0066 0.0054 0.0055 0.0039 0.0039 0,0026 0. 0028 0.0018 0.0022 0.0016 0.0012
0.1050 0.0957 0.0875 0.0795 0.0726 0.0659 0.0601 0.0548 0.0498
0.0457
0.0387 0.0327 0.0276 0.0237 0.0203 0.0177 0.0152 0.0131 0.0114 0.0099 0.0085 0.0074 0.0056 0.0043 0.0035 0.0028
TABLE VI1 CURVEFITTINGTO EQUATION (23)
OF
(1962)
BAKERAND
2
1
3.12
4.24
3
3.99
4.04 1.94
Standard error Absolute error
6.39 2.62
4.48 1.41
4.65 0.97
4
3.50 5.85 1.07
b~ is Zero
5.27 3.39
3.80 1.02
610.0 582.0
bl and br Are Zero
Standard error Absolute error
17.1 13.2
29.8 27.7
58.3 55.5
+ A eEA
+B
EXY
E EA’= +B’
+A’
E
+ A’
must be entertained.
ACKKOWLEDGEMEXTS The author wishes to thank Profs. H. R. Mahler and V. J. Shiner, Jr., for many helpful discussions and for their advice in the preparation of the manuscript, and Prof. E. L. King for his suggestions. Mrs. Margaret Hanes provided capable technical assistance in many phases of this investigation.
REFERENCES Abeles, R. H., Frisell, W., and Mackensie, C. (1960), J . Biol. Chem. 236,835. Alberty, R. (1958), J . Am. Chem. SOC.80, 1777. Baker, R. H., Jr., and Mahler, H. R. (1962), Biochemistry 1, 35.
10, 447.
All Constants Uaed
Standard error Absolute error
E
Bonnichsen, R. K., and Brink, N. G. (1958), in Methoda of Enzymology, vol. I, Colowick, S. P., and Kaplan, N. O., ed., New York, Academic Press, Inc., p. 495. MAHLER Boyer, P. D., and Theorell, H. (1956), Acta Chem Scand.
Exueriment
Obsewed error
obligatory manner. The mechanism probably involves one or more ternary complexes, and the likelihood that the enzyme forms complexes other than those involved in the simple, linear scheme
60.9 57.9
errors plus negative errors). I n all of the examples except (3), the standard error for the points computed with the use of all the coefficients is less than that found when any of these are set equal to zero. Even in this one case, as in all of the others, the absolute error, a measure of the over-all fit, is smaller. I n case (4), the extreme error in the fit with bl set equal to zero is caused by a discontinuity of the computed points. Therefore, it can be surmised that the best fit is obtained by the use of all the coefficients. Hence, mechanisms IIa, b, and c are considered the most likely.6
DISCUSSION From the foregoing we conclude that under the particular conditions described in this paper the 6 The constants derived have not been presented, as they are simple fitting parameters with the possibility of being complicated groups of rates. No attempt has been made in the treatment of these numbers to solve for the individual rates involved.
Collins, C. J. (1960), Ann. 1%’. Y . Acad. Sci. 84, 603. Dalziel, K. (1958), Acta Chem. Scand. 12, 459. Ehrenberg, A., and Dalziel, K. (1958), Acta Chem. Scand. 11, 465.
Englard, S. (1960), Ann. N.Y. Acad. Sci. 84, 695. Horecker, B. L., and Kornberg, -4. (1948), J . Biol. Chem. 1 Y6,385. Kornberg, A., and Pricer, W. E., Jr. (1953), Bwchm. Prep. 3, 20. Levy, H. R., and Vennesland, B. (1957), J . Biol. Chem. 288, 85. Mahler, H. R., and Douglas, J. (1957), J . Am. Chem. SOC. Y9. 1159.
Melander, L. (1960), Isotope Effects on Reaction Rates, New York, The Ronald Press. Memitt, A. D., and Tomkins, G. M. (1959), J. Biol. Chem. 234, 2780. Rafter, G. W., and Colowick, S. P. (1058), in Methods of Enzymology, Colowick, S. P., and Kaplan, N. O., ed.. vol. 111, New York, Academic Press, Inc., p. 887. Rieder, S. V., and Rose, I. A. (19591, J . Biol. Chem. 234, 1007.
Roii:I. A. (1961), J . Biol. Chem. $36, 603. Shiner, V. J., Jr. (1952), J . Am. Chem. SOC.Y6,2929. Shiner, V. J., Jr., Mahler, H. R., Baker, R. H., Jr., and Hiatt, R. R. (1960), Ann. N . Y . Acad. Sei. 84, 583. Streitwieser, A., Jr. (1960), Ann. N . Y . Acud. Sca. 84, 576. Theorell, H. (1956), in Currents in Biochemical Research, Green, D. E., ed., New York-London, Interscience, p. 293.
Theorell, H. (1958), in Advances in Enzymology, vol. 20, Nord, F. F., ed., New York-London, Interscience, p. 31. Theorell, H. (1961), Fed. Proc. 20, in press. Theorell, H., and Bonnichsen, R. K. (1951), Acta Chem. Scand. 6, 1105. Theorell, H., and Chance, B. (1951), Acta Chem. Scand. 6, 1127.
Theorell, H., Nygaard, A. P., and Bonnichsen, R. K. (1955), Acta Chem. Scand. 9, 1148. Theorell, H., and Winer, A. D. (1959), Arch. Biochem. Biophys. 83,291. Vallee, B. L., and Coombs, J. L. (1959), J . BioZ. Chem. $34, 2615.
VOl. 1, NO.1, January 1962
MECHANISM OF LIVER ALCOHOL DEHYDROGENASE REACTION.
Vallee, B. L., Williams, R. J. P., and Hoch, F. L. (1959), J . Bwl. Chem. 234, 2621. Van Eys, I., Stolzenbach, F. E., Sherwood, L., and Kaplan, N. 0. (1958), Biochim. Biophys. Acta 27, 63. Vennesland, B. (1958), Fed. Proc. 17, 1150.
IV
47
Vennealand, B., and Westheimer, F. H. (1954), in The Mechanism of Enzyme Action, McElroy, D., and Glass, B., ed., Baltimore, The Johns Hopkins Press p. 357. Wiberg, K. B. (1955), Chem. Revs. 65, 713. Winer, A. D. (1958), Acta Chem. Scund. 12, 1605.
Studies on the Mechanism of Enzyme-Catalyzed Oxidation Reduction Reactions. IV. A Proposed Mechanism for the Over-all Reaction Catalyzed by Liver Alcohol Dehydrogenase* H. R. MAHLER, R. H. BAKER, JR.,t
AND
V. J. SHINER, JR.
From the Department of Chemistry, Indiana University, Bloomington, Indiana Received August 11 , 1981
The kinetic data for liver alcohol dehydrogenase acting on acetaldehyde-ethanol obtained in the preceding paper are analyzed in terms of a variety of possible modifications of the mechanisms proposed by Theorell and Chance (1951) and ourselves. It is concluded that the most likely mechanism is one entailing the formation of both unreactive and reactive complexes between both enzyme and DPN and enzyme and DPNH. The kinetic and equilibrium dissociation constants obtained on the basis of the proposed mechanism are shown to be in good agreement with values determined independently for &q.over-d by two techniques by ourselves and by direct measurement by Backlin; with values for the dissociation constant of the enzyme-coenzyme complexes obtained by direct measurement by Theorell and Winer; and with values for the kinetic constants obtained by Theorell et al. The isotope effects for the binding constant for DPNH and for the rate constant of the hydrogen transfer step proper have been determined by comparing DPNH with a-DPND in the reactions catalyzed by both liver and yeast alcohol dehydrogenase, and found to be of the order of 1.3 for the equilibrium constant and 2-3 for the rate constant. A complete kinetic analysis is also presented for the reaction in the presence of the inhibitor o-phenanthroline. It is concluded, in agreement with Vallee and co-workers, that both DPNH and DPN bind a t a site also capable of interacting with the inhibitor and therefore probably a t the enzyme-bound Zn++. Ethanol binds a t a site not identical with this, but related to it or closely adjacent, while acetaldehyde seemingly does not bind a t the same site (or in the same manner) as ethanol. I n the preceding paper (Baker, 1962), the first is the formation of isomeric forms of the binary mechanism of the liver alcohol dehydrogenase complexes between enzyme and the leading subsystem has been investigated by a study of its strate (Peller and Alberty, 1959). As an example, kinetics. It became apparent that the most this particular modification of the Theorell-Chance satisfactory description involved some type of mechanism (1951), as the simplest example of a compulsory binding mechanism, but that none compulsory binding type, is shown in Equation (1) of the examples in this group (homeomorphs) described by Dalziel (1957) could account for some of the characteristics of the reaction, notably the fact that it did not obey the “Dalziel criteria.” ki ko Therefore some modification or perturbation has E Y eEA’ 5 E + A’ ( 1 ) to be introduced. I n general, with mechanisms ks kio which do not postulate alternate pathways (forks) between reactants and products, only two classes A and A’ correspond to DPNH and DPN in the of perturbations exist which can alter the Dalziel present case. relationships in the observed direction.’ The ks * The investigations described in this paper were supEX + B e B’ EY ported by a research grant ONR 908(12) from the Office of Naval Research to Indiana University. t National Science Foundation Co-operative Predoctoral Fellow 1959-60. Present address, Department of Chemistry, University of Wisconsin, Madison. The material presented in this and the preceding pa ers is taken in large part from a dissertation of R. H. Raier, Jr., submitted to the Graduate. School of Indiana University in partial fulfillment of the requirement for the Ph.D. degree. $ Contribution No. 1030. 1 This becomes readily apparent if the schematic method of King and Altman (1956) is applied.
+
E/VO=
’pu
+
+
ks e/@) rnv~/(A)(B) ( l a )
+
where (A) and (B) are the initial concentrations of the two substrates and , + ks * = k--, kks
(,12
= kn --__
kikh
(lb)
